What is mattwood.fyi?

A live, structured feed of riffs and links from Matt Wood, Chief AI & Technology Officer at AWS — readable by people and queryable by agents as a typed knowledge graph.

Is there an API and is it free?

Yes. A public, read-only query API is available with no authentication and no API keys. See https://mattwood.fyi/openapi.json and https://mattwood.fyi/docs.

How should an AI agent use this site?

Fetch https://mattwood.fyi/feed.json for the full structured feed, or query the API for keyword/semantic search and the connection graph. See https://mattwood.fyi/agents/ for when-to-use guidance.

Code Execution with MCP: Building Efficient AI Agents

Connections

Related to

AX: Agentic Task Execution Platform AX agentic task execution platform represents the broader infrastructure context where MCP code execution patterns would be implemented as a core execution primitive
The Plunging Price of Thought Reducing token consumption via MCP code execution is a direct response to the economics discussed in 'The Plunging Price of Thought' - optimizing agent efficiency as inference costs shift
Formal Methods for Controlling AI Agents MCP as a structured protocol for AI-tool interaction raises the same formal control concerns addressed in formal methods for controlling AI agents - both grapple with bounded, verifiable agent behavior

Supports

Unreal Agent: Cost-Efficient AI Agent Harness Both focus on cost-efficient AI agent execution; MCP code execution approach directly enables the kind of token-efficient agent harness architecture described in Unreal Agent
Unreal Agent: Cost-Efficient AI Agent Harness Both focus on cost-efficient AI agent architectures; MCP code execution approach directly reduces token consumption, aligning with Unreal Agent's cost-efficiency goals
SoL-Pi: Recursively Scaling Auto-Research Loops for Efficient Agent Harness Both papers address efficient agent harness design; MCP code execution reducing context window usage directly contributes to the scaling efficiency goals described in SoL-Pi's auto-research loops
The Plunging Price of Thought Token consumption reduction via MCP code execution is a direct response to the economics of thought described in 'Plunging Price of Thought' — efficiency gains enable more agentic compute per dollar
Strands Harness: Frontier Performance with 28% Lower Token Cost MCP code execution reducing token consumption aligns directly with Strands Harness achieving 28% lower token cost — both are architectural approaches to agent efficiency
SoL-Pi: Recursively Scaling Auto-Research Loops for Efficient Agent Harness SoL-Pi's efficient agent harness loops would benefit from MCP's code execution model, as reducing context window pressure is critical for recursive/scaling agent architectures
Strands Harness: Frontier Performance with 28% Lower Token Cost MCP code execution as a mechanism for reducing token costs directly supports Strands Harness's 28% lower token cost methodology - both target the same optimization problem

Develops into

AX: Agentic Task Execution Platform MCP-based code execution for tool interaction represents a foundational pattern that agentic task execution platforms like AX would build upon for orchestrating complex workflows
Frontier engineering: Working with AI agents MCP code execution is a concrete technical pattern within the broader frontier engineering with AI agents space, offering practitioners a specific methodology discussed abstractly there

Type:	link
First published:	Friday, 25 September 2026 at 22:50 UTC
Location:	https://www.anthropic.com/engineering/code-execution-with-mcp
URL:	https://mattwood.fyi/i/a2e5eb08

I'm Matt Wood, and this is For Your Information. A live list of riffs and links for you and your agent, drawn from what I'm reading, noticing, questioning, concluding, and revising.

Connections